Fast Screening Method of Biological Samples Based on Needle Stochastic Sensors for Early Detection of Gastric Cancer

Early detection of cancer is essential for saving the life of patients. Needle stochastic sensors were proposed as new tools for the fast screening of biological samples. The design of the stochastic sensors was based on the immobilization of protoporphyrin IX in pastes of S-doped graphene. The selected biomarkers were CEA and CA19-9. High sensitivities, and wide linear concentration ranges as well as low limits of quantification were achieved. The proposed sensors were validated; student t-test shown that the proposed needle stochastic sensors can be reliably used for the pattern recognition and quantification of CEA and CA19-9 in whole blood, gastric tumor tissue, saliva, and urine.


Introduction
Patients with gastric disease have no undeniable side effects, and a large number of them are in the center or propelled stages while accepting investigation in an emergency clinic, in this manner passing up on the best open door for treatment [1,2]. Gastroscopy joined with obsessive conclusion is the best quality level for analysis of gastric malignant growth, however poor resilience to gastroscopy, high monetary expenses and low worthiness among patients make it unsatisfactory for fundamental screening [3,4]. Metabolites straightforwardly created by tumor cells and discharged into body liquids or tissues, or expert deuced by ordinary cells when battling against tumors fill in as tumor markers, and normally exist as antigens, compounds or hormones [5]. In the typical condition, such markers show no presence or amazingly low substance in the body. Their quality or changes in amount can mirror the event and movement of tumors, with points of interest of advantageous detection, solid adequacy and low costs, which improves the location pace of essential tumors, screen tumor repeat or metastasis, and assess the anticipation impact [6].
The forecast of early gastric malignant growth is good after revolutionary gastrectomy, with a 5-year by and large endurance rate surpasses 97% [7]. An assortment of elements have been perceived as prognostic elements for early gastric malignant growth, including tumor size, separation status, tumor profundity, LNM and vessel association [8]. Also, tumor markers including carcinoembryonic antigen (CEA) [9], cancer-related antigen 19-9 (CA19-9) [10], and AFP [11] were exhibited to be prognostic variables for gastric malignancy. CEA and CA19-9 are some of the biomarkers that are present in gastric cancer. When these biomarkers are found in the human body in an increased concentration, they indicate the presence of gastric cancer. CA 19-9 is a well-known biomarker of digestive tract malignancies. Although tumor biomarkers are an invaluable asset to medical practice, their role in screening, diagnosis and oncologic treatment remains poorly standardized [12,13]. In gastric cancer, the most used tumor markers are: CEA and CA 19-9 [12].
Stochastic sensors have been used to date for the molecular recognition and quantification of several biomarkers used in the diagnosis of different illnesses such as diabetes, cancer, and obesity [20][21][22]. Their main advantages are the possibility to perform a reliable qualitative and quantitative analysis of various biomarkers at very low concentrations in complex samples, no treatment (sampling) was needed on the biological samples before performing the analysis, and the results of the molecular recognition and quantification were not influenced by the complexity of the matrix from which the biomarkers were determined.
In this study, protoporphyrin IX (PIX) was used as the recognition molecule for the design of the needle stochastic sensors for the simultaneous recognition and quantification of CEA and CA19-9 in biological samples such as: whole blood, saliva, urine, and tissue samples.

Materials and reagents
CA19-9, CEA, monosodium phosphate and disodium phosphate were purchased from Sigma Aldrich (Milwaukee, USA); paraffin oil was purchased from Fluka (Buchs, Switzerland). Monosodium phosphate and disodium phosphate were used for the preparation of phosphate buffer, pH=7.5. Deionized water obtained from a Millipore Direct-Q 3 System was used for the preparation of all solutions. Dimethylsulfoxide (DMSO) and sodium sulfide (Na2S) were purchased from Merk (Germany); sulfur was acquired from Remed Prodimpex (Romania).

Apparatus and methods
For all measurements, a potentiostat/galvanostat AUTOLAB/PGSTAT 302 (Methrom, Utrecht, The Netherlands), connected to a personal computer with a GPES software installed was used. An electrochemical cell, containing a three electrode system was also employed. The three electrode system is made out of: the counter electrode (platinum wire), the working electrode (proposed needle stochastic sensors) and the reference electrode (Ag/AgCl electrode).C/H/S contents in the prepared samples were obtained using elemental analysis (FLASH EA 1112, Italy) and EDX. Samples were morphologically characterized by TEM (H-7650 120 kV Automatic TEM, Hitachi, Japan). The structural data were obtained from XRD patterns (Bruker D8 Advance diffractometer, Germany).

Preparation of S-dop-1
The preparation of S-dop-1 was conducted by a solvothermal method using DMSO as reducing/doping agent. Briefly, 150 mg of GO, previously synthesized [23] were dispersed in 55 mL DMSO and kept at 180°C for 18 h. The obtained material was washed with ethanol and water, to remove the unreacted DMSO and the unpleasant odored by-products, then dried by lyophilization.

Preparation of S-dop-Gr-1-S
For the preparation of S-dop-Gr-1S we used 240 mg GO dispersed in 80 mL H2O then 240 mg of Na2S were added; the mixture was put in an autoclave and kept in oven for 10 h at 200°C. After cooling to room temperature the sample was washed with double-distilled water, filtered and dried by lyophilization.

Preparation of Gr-S5-TT
700 mg GO were sonicated for 25 min in 140 mL toluen, then 500 mg S was added and sonicated another 30 min. The obtained dispersion was kept for 12 h at 160°C. The prepared material was washed with toluene, filtered, and dried by lyophilization. For further purification the sample was thermally treated, 15 min at 450°C.

Design of the needlestochastic sensors
Three types of S-doped graphene powders were mixed with parrafin oil, until three homogeneous pastes were obtained. A 10 -3 mol L -1 protoporphyrin IXsolution was added to each paste (100μL solution:100mg graphene paste). Each paste was placed in a non-conducting plastic tube with internal diameter of 100μm.The conection between the paste and the external circuit was made using a silver wire. Before and after each measurement, the sensors were cleaned with deionized water, and dryed. When not in use, the three stochastic sensors were kept in the refrigerator, in a dark box.

Stochastic mode
The stochastic mode was used for all measurements. The principle of the stochastic sensors is based on the channel conductivity. Chronoamperometry was selected and a constant potential of 125 mV was applied. The parameters read in the diagrams were toff and ton. The biomarkers: CEA and CA19-9 were identified in the diagrams (Figures 1-4) based on their signatures(toff values), while reading the corresponding ton value in between two toff values, the determination of the concentration of the biomarker was possible by inserting the ton values in the equation: where a is the intercept, b is the slope and c is the concentration of the biomarker. The equations of calibration for each sensor, and each biomarker were statistically calculated using the ton values obtained after measurements of the standard solutions (of different concentrations) of CEA, and CA 19-9.

Samples
Whole blood, saliva, urine and tissue samples were recieved from the County Emergency Hospital from Targu-Mures. Samples from confirmed patients with gastric cancer were collected according to the procedures specified in the Ethics committee approval number 32647/2018 awarded by the County Emergency Hospital from Targu-Mures. Written consent was obtained from all patients. No sample pretreatment was performed.

Characterization of the S-Gr samples
According to the elemental analysis, the samples prepared with sulfur (Gr-S5-TT) and sodium sulfide (S-dop-Gr-1-S) as doping agents have low content of sulfur whereas the other sample, S-dop-1, prepared in DMSO, has a higher content of sulfur, 5.42 wt% (Table 1). In Figure 5 a-c are shown representative TEM images of the prepared samples. It can be observed the wrinkled graphene surface with narrow dark lines which represents the edges of the folded graphene.

Response characteristics of the stochastic sensors
The process of current development in stochastic sensors comprised two phases [22]:the first phase is known as the pattern recognition phase; the biomarker is extracted from the solution into the membrane-solution of the interface and blocks the channel; because of this, the intensity of the current decreases to 0 value for a certain period of time, which is known as toff and it represents the signature of the biomarker; its value is used for the qualitative analysis of the biomarker; the second phase is known as the binding phase and it takes place when the analyte interacts with the wall channel, and the redox processes take place. During the second phase, the following equations of binding take place: where Ch represents the channel and i represents the interface. The quantitative parameter, also known as ton, represents the equilibrium time needed for the interaction between the biomarker and the wall channel/redox processes. Table 2 shown the response characteristics of the proposed needle stochastic sensors. The highest sensitivity for the determination of CEA was obtained when the needle stochastic sensor based on Gr-S5-TT was used; while for the assay of CA19-9, the highest sensitivity was obtained when the sensor based on Sdop1 was used. The linear concentration range for the assay of CEAwas obtained when the sensor based on S-dop-Gr-1-S was used while for the assay of CA-19-9 the widest linear concentration range was obtained when the sensor based on Gr-S5-TT was used. The lowest limit of quantification for https://doi.org/10.37358/RC.21.4.8452 the determination of CEA was obtained using the sensor based on S-dop-Gr-1-S while for the assay of CA-19-9 the lowest limit of quantification was obtained when the sensor based on Gr-S5-TT was used.

Selectivity
Different toff values, also known as the signatures of the biomarkers of interest, were recorded for CEA and CA19-9, using the same sensors (Table 2). When checking p53, maspin, D-glutamine as possible interference, for all sensors, signatures higher than 2.4s were obtained, proving that the proposed needle stochastic sensors are selective.

Stability
The needle stochastic sensors have been used continuously for one year for measurements. Comparing the sensitivities obtained every day, each with the other, a relative standard deviation value lower than 1.00% was obtained. Accordingly, the proposed needle sensors are highly stable.

Determination of CEA and CA 19-9 using a fast screening method of biological samples
Biological samples: urine, whole blood, saliva and gastric tissue samples, were collected from confirmed patients with gastric cancer. All samples were analyzed without any pretreatment. Molecular recognition based on the signatures of CEA and CA19-9 was first performed, followed by thereading of the tonvalues, and the determination of the concentrations of CEA and CA19-9 accordingly with the stochastic mode described above.
The obtained results for the determination of CEA and CA19-9 in urine, whole blood, saliva and tissue samples from patients confirmed with cancer are shown in Tables 3-6. A good correlation between the obtained results using the proposed stochastic sensors were obtained for whole blood, gastric tissue samples, saliva, and urine samples, indicating that the proposed needle stochastic sensors may be reliably used for the screening tests used for the molecular recognition and determination of CEA and CA19-9 in biological samples. Student t-test were also performed at 99.99% level; all calculated t values were lower than 4.3 which is the tabulated value; these results also certified that the proposed needle stochastic sensors can be reliably used for pattern recognition and quantification in fast screening tests of whole blood, gastric tumor tissue, saliva, and urine samples.